Analysis of Musical and Voice Skills in Children and Adult Cochlear Implant Users

METHODOLOGY

Design Cross-sectional study in which both implanted patients and control subjects without hearing loss were tested only once.

Scope and study subjects Prior to the inclusion of subjects in the study, the approval of the Ethics Committee of the Hospital Universitario La Paz (HULP) will be required.

STUDY GROUP (Implanted group) Adult patients (≥18 years) with postlingual hearing loss and children (between 6 and 17 years) with prelingual or perilingual hearing loss who have received a CI before the age of 3 years.

Patients who have received a CI from 2000 to January 2023 will be included as potential candidates for participation. Currently 76 subjects aged 6-17 years implanted before the age of 3 years and 231 post-lingual adults are identified in the database, to which post-lingual adults implanted in the next 2 years should be added, so that after one year of follow up they can participate in the study.

Inclusion criteria:

• Minimum age of 6 years.
• Patients aged 6-17 years must have been implanted before the age of 3 years (to ensure adequate CI performance).
• Minimum 12 months experience with the device
• Indications: unilateral (one CI), bilateral (two CIs), bimodal (one CI in one ear and a hearing aid in the other)
• At least 80% of electrodes active at the time of assessment.
• Fluency in Spanish
• Signed and dated informed consent prior to starting any study procedure.

Exclusion criteria:

• Not meet inclusion criteria.
• Prelingual or perilingual adults.
• Disability in addition to hearing impairment (e.g., mental disability). Routine visits in the HULP or at their CI fitting sessions will be used as an opportunity to contact patients and propose their participation.

CONTROL GROUP The control group will be composed of subjects over 6 years of age without hearing impairment, without a disability that prevents them from taking the tests and who have socio-demographic characteristics similar to those of the study group. They will be included from the ENT Department where they will have come either with another type of medical condition other than hearing loss, or accompanying patients. Hospital La Paz staff (or family members) who wish to carry out the tests may also be included. They must be fluent in Spanish to ensure that they understand the study and its purpose. To confirm that the control group does not have hearing impairment, they will undergo tonal audiometry in both ears, whose tonal threshold should be ≤30dB in the conversational frequencies of 500, 1000 and 2000Hz.

Socio-demographic and clinical variables Study group: affiliation data (current age, age at implantation, sex, education level), hearing loss data (aetiology, duration of hearing loss, type and degree of hearing loss, worst ear), surgery data (type of implant and electrode array used, complete or incomplete insertion, side implanted, date of intervention), postoperative complications. These data will be collected from the patients' medical records. In addition, fititing data (number of active electrodes, processor model) and audiological tests with the CI will be collected.

Control group: age, sex, associated pathology(ies), level of education.

Outcome variables

1. Scoring of the music-related quality of life questionnaire (MuRQoL), the Munich music questionnaire (MuMu) and the questionnaire developed for subjects aged 6-17 years (see description).
2. Validation of the music-related quality of life questionnaire (MuRQoL) in adult population.
3. Score obtained in the "discovery" level with the Meludia music tool (see description).
4. Parameters to assess voice quality: Spoken voice: Vocal inability index, Maximum phonation time, Fundamental voice frequency, Jitter, Shimmer, Harmonics to noise ratio (HNR), Smoothed cepstral peak (CPPS) and Spectrogram according to Yanagihara's classification. Sung voice: Pitch (relative to the note emitted by a keyboard), CPPS and Singing power ratio (SPR).
5. Impact of demographic and audiological factors on the results of the music perception questionnaires as well as on the results of the Meludia tool and the acoustic analysis of the implanted patients and comparison with the normal hearing control group.

Description of the measurement instruments

1. - MuMu Questionnaire: The Munich Music Questionnaire was especially developed for the adult CI users population with postlingual hearing loss. It includes questions that address past and present activities related to the habit of listening to music. We will use the abbreviated version which contains 9 questions on music listening habits in relation to various musical styles, different instruments and participation in musical activities at different times in the life of implanted patients. It offers many response options, from the "yes or no" format to rating scales and multiple choice questions with the possibility to tick as many options as necessary (e.g. "How often do you listen and/or have you listened to music", on a scale of 0-10; "Why do you listen to music", with the options of "for pleasure", "professional reasons", "emotional satisfaction", "to relax"). It is translated and validated in Spanish (developed by S.J. Brockmeier and adapted by Ana Fuster. Info: https://www.medel.com/support/rehab/rehabilitation-downloads).
2. MuRQoL Questionnaire: The concept of music related quality of life (MuRQoL) was initially defined as the impact of listening skills on the quality of life of implanted patients. To work on this concept, groups of adult CI users created the questions, thus ensuring content validity, differing from previous studies in that users were not involved in the initial question generation process. The MuRQoL questionnaire addresses novel aspects of musical experience, such as the ability to hear whether one is singing at the right pitch. It uses a 5-point Likert-type scale or satisfaction scale, which includes a range of opinions allowing for a greater degree of specificity than a yes/no question. Five of the 18 questions included are novel among music questionnaires designed for adult CI users, (question 4 - "can you recognise words in songs?", 6, 9, 10 and 15 - "do you listen to music while travelling, e.g. in the car?"). This is important for assessing musical experiences that have not been measured by previous questionnaires. On the other hand, several questions are similar to those of other questionnaires, e.g. question 18 is also reflected in the MuMu test ("Do you sing, play a musical instrument or whistle?"). The questionnaire consists of two parts with the same 18 questions each, divided into musical perception and musical engagement. Part I of the questionnaire asks about musical experiences, e.g. "Can you differentiate between different rhythms in music? Part II asks about the importance of music listening skills, attitudes towards music and musical activities for the implant recipient, e.g. "How important is it for you to be able to distinguish different rhythms in music?" Dritsakis G, van Besouw RM, Kitterick P, Verschuur CA. A Music-Related Quality of Life Measure to Guide Music Rehabilitation for Adult Cochlear Implant Users. American journal of audiology 2017;26:268-282.
3. Questionnaire for subjects aged 6-17 years: After reviewing the existing literature, we did not find any questionnaire suitable both in language and in the form of scores for subjects aged 6-17 years. Therefore, we have made a selection of questions that have already been used by other authors and that encompass aspects of musical interests, musical profile, role of music and music-related activities [Pacheco et al., 2020; Vannatta-Hall, 2010; Waaramaa et al., 2018] (See Annex).Pacheco L, Miguel JHS, Gil D. Musical stimulation proposal for hearing impaired children: case reports. CoDAS 2020;32:e20190139; Vannatta-Hall JE: Music education in early childhood teacher education: the impact of a music methods course on pre-service teachers' perceived confidence and competence to teach music. University of Illinois at Urbana-Champaign, 2010; Waaramaa T, Kukkonen T, Mykkänen S, Geneid A. Vocal Emotion Identification by Children Using Cochlear Implants, Relations to Voice Quality, and Musical Interests. Journal of speech, language, and hearing research : JSLHR 2018;61:973-985.
4. Meludia: The music tool Meludia (www.meludia.com), available in Spanish, uses interactive music listening exercises of progressive difficulty for both children and adults. It is based on listening and responding. Meludia poses questions using sounds and colourful graphic designs, which after listening carefully (patients controlled directly and those implanted through the connection cable that allows direct listening through the audio processor) are selected with the PC mouse by clicking on the button that best corresponds to what has just been heard. Different categories are evaluated: Rhythm (indicate how many percussion beats are heard, from 1 to 8), Melody (indicate whether the melody heard is ascending or descending), Density (indicate how many sounds are heard simultaneously, one or several), Stability (indicate whether the sound heard is chaotic or not; that is, whether it gives a feeling of stability or instability) and Spatialisation (indicate whether the second note has a higher or lower pitch than the first one). The tool consists of 4 modules: Discovery, Intermediate, Advanced and Expert. In this study we will initially work with the Discovery module, which consists of 25 exercises (5 for each of the 5 categories indicated above) (see appendix). Scores (between 1 and 3) will be collected for each of the exercises, as well as whether or not each category is completed. A maximum of 4 attempts will be allowed in each of the exercises. If after the 4 attempts the subject is not able to obtain at least a score of 1, the test shall stop in that category and move on to the next category. The test shall start with the "rhythm" category, which is considered the easiest category.
5. Acoustic analysis of the voice: the Praat programme will be used (developed by Paul Boersma and David Weenink of the University of Amsterdam), which is the most widely used in publications related to the subject 6 and whose values have been compared with other programmes validated in Spanish 7. The sound signal will be collected with a Saramonic International microphone (3.5 mm) and a spoken voice recording will be made according to the protocol of the European Society of Laryngology 8:

• Vowel /a/ held for at least 3 seconds.
• Phonetically balanced text ("Platero y yo" by Juan Ramón Jiménez).
• Voice recording sung with the first phrase of the popular song "Cumpleaños feliz" prolonging the last vowel /i/ for at least 3 seconds. The parameters to be analysed in order to evaluate the quality of the voice are the following: Spoken voice: Vocal incapacity index, Maximum phonation time, Fundamental frequency of the voice, Jitter, Shimmer, Harmonics to noise ratio (HNR), Smoothed cepstral peak (CPPS) and Spectrogram according to Yanagihara's classification. Sung voice: Pitch (with respect to the note emitted by a keyboard), CPPS and Singing power ratio (SPR).

1. - Description of the audiological tests: the study group will undergo tonal and speech audiometry in free field to obtain objective values of their CI performance. This will be carried out in a double-walled acoustically insulated booth and the subjects will be seated 1 m away from the speakers at 0º azimuth. If a patient has better hearing in the non-implanted ear, this ear will be masked during the test thus eliminating the perception of the "good ear" and assessing CI performance only, and if the patient is wearing a hearing aid it will be removed for the test. For tonal audiometry the frequencies of 250, 500, 1000, 2000, 4000 and 6000Hz will be evaluated and for statistical purposes the PTA4 value (average thresholds at 500,1000, 2000 and 4000Hz) will be used. Speech perception will be evaluated with the % of disyllables both in quiet and in noise, without lip reading, at 65dB SPL, and with a signal to noise ratio of 10dB SPL below the signal; the recorded tracks of Cardenas and Marrero will be used.

Procedure for completion of the questionnaires and performance of the musical and voice test *Study group: After receiving information about the study, both oral and written, and signing the informed consent form, patients will fill in the MuMu and MuRQoL questionnaires (adults) and the questionnaire developed for subjects aged 6-17 years with the help of a member of the research team (children will also be assisted by their parents or guardians).

The musical test with the Meludia tool, which we estimate to last approximately 1 hour, will be carried out after completing the questionnaires, or else they will be scheduled for another day within the following 30 days. The audio input of the CI processor will be connected to the audio output of the laptop computer, in order to completely isolate the signal from external noise. Once all the Meludia exercises have been completed, the voice will be analysed with the Praat programme.

*Control group: normal hearing patients will receive oral and written information about the study and after signing the informed consent they will fill in the MuRQoL questionnaire if they are over 18 years old or the one developed for subjects between 6-17 years old. They will then be assessed with the Meludia tool and subsequently with the Praat programme.

Both the questionnaires and the test with the Meludia tool will be administered to each patient individually in a quiet environment. If the youngest children do not manage to complete all the exercises in a single session, they will be scheduled for the following days, always within the next 30 days, taking advantage of one of their routine check-ups.

Given the current pandemic situation due to SARS-COV2, in order to minimise the time it takes for both the study group and the controls to visit the hospital, a self-completing version of the questionnaires would be created so that they can be filled in online. Assessment with Meludia can also be performed remotely via Webex, Zoom or Teams; in this case, implanted patients would have the cable connecting the processor(s) to the audio output of the PC delivered to their homes. In the same way as the voice analysis, which, after sending the corresponding microphone to the subjects, could be carried out remotely.

Since the study tests (Meludia music tool, questionnaires and voice analysis) will only be tested once on each subject, there is no need to follow up the groups. While all implantees have an annual check-up throughout their lifetime, no changes in performance are expected unless there is an accident or device failure. Therefore, the results of the study are not expected to change over time. The control group, on the other hand, is made up of normal hearing subjects, with the ability to perform the tests correctly, i.e. without a disability that prevents them from doing so (e.g. mental impairment) and this ability should not vary substantially over time. As there is no follow-up on the subjects (the "measures" are only performed once in time) there will be no losses in the study.

Translation of the MuRQoL questionnaire Initially, the questionnaire will be translated from English to Spanish by a suitably qualified person. This will be followed by translation from Spanish to English (back-translation) by another qualified person from outside the clinical study. This will ensure that the Spanish version faithfully reproduces the original English version.

Statistical analysis Quantitative variables will be expressed as mean, standard deviation and range, and qualitative variables as absolute frequencies and percentages. For the analysis of the results of the MuMu music perception test before and after the CI, the Wilcoxon test for paired data will be used. To assess the influence of socio-demographic and clinical variables on the results of music perception after the hearing implant, comparisons of means between groups will be made using non-parametric methods (Kruskall-Wallis for more than two groups and Mann-Whitney for two groups) or parametric methods (ANOVA for more than two groups and Student's t-test for two groups), depending on the sample sizes, the homogeneity of the data and the adjustment to the normality of the distribution. Spearman's correlation coefficient will be used to analyse the relationship between verbal discrimination after CI and the measurement of music perception and enjoyment. All analyses will be performed at a two-tailed p<0.05 level of statistical significance. The statistical package SPSS v.24 will be used for the proposed analyses.

For the validation of the MuRQoL questionnaire, the method of Rust \& Golombok, 2000 will be used. Reliability and validity will be taken into account. Reliability is assessed through consistency (assessed through Cronbach's alpha coefficient), temporal stability (test-retest reliability, calculated with the intraclass correlation coefficient) and inter-observer agreement (analysed through the percentage of agreement and the Kappa index), and validity through content validity (qualitative assessments to be made by expert researchers), construct validity (factor analysis and multitrait-multimethod matrix) and criterion validity (relation of each subject's score to a Gold Standard, in this case to the scores obtained in the original item [Dritsakis et al. , 2017]).

Dritsakis G, van Besouw RM, Kitterick P, Verschuur CA. A Music-Related Quality of Life Measure to Guide Music Rehabilitation for Adult Cochlear Implant Users. American journal of audiology 2017;26:268-282.

Regarding sample size determination, this is the first study with the Meludia music tool in a child population, so it is not possible to know with certainty the sample size needed to find significant differences between the study group and the control group. But based on a recent study published by our group where there is an implanted population and their control peers ("matched" by gender, age, educational level...), the number of patients in the study group and 24 controls was set at 24 (Mertens G., et al. 2020. Cognitive Improvement After Cochlear Implantation in Older Adults With Severe or Profound Hearing Impairment: A Prospective, Longitudinal, Controlled, Multicenter Study. Ear and hearing), with 75% power to detect a difference of 14 standard deviations (= Cohen's d; Claes, A. J., et al. 2018. Cognitive performance of severely hearing-impaired older adults before and after cochlear implantation: Preliminary results of a prospective, longitudinal cohort study using the RBANS-H. Otol Neurotol, 39, e765-e773).

Limitations of the study The retrospective design for the selection of patients is not an obstacle since the data collection, the completion of the questionnaires and the musical test with Meludia are evaluated prospectively, allowing us to control possible biases related to this type of design, such as the loss of information.

One of the possible limitations we might encounter is the refusal of some patients to stay longer than necessary in a hospital room. This could be solved by offering them the possibility of doing the exercises online, with an internet connection at home, and we could guide them through the exercises.